Blame for report.py

import matplotlib.pyplot as plt
from fpdf import FPDF
import pandas as pd
import uuid

from langchain_core.output_parsers import StrOutputParser
from langchain_core.messages import HumanMessage
from langchain_ollama import ChatOllama

import matplotlib.pyplot as plt

import base64
from io import BytesIO
from PIL import Image

from datetime import datetime
import random

class ReportGenerator(FPDF):
    def __init__(self, start_date, end_date):
        super().__init__(unit='cm')

        self.start_date = start_date
        self.end_date = end_date

        self.model = ChatOllama(model="llava-phi3:3.8b", temperature=0)
        self.chain = self.prompt_func | self.model | StrOutputParser()


    def generate_months(self):
        # Convert input strings to datetime objects
        start = datetime.strptime(self.start_date, '%b %Y')
        end = datetime.strptime(self.end_date, '%b %Y')

        # Generate the list of month-year strings
        months = []
        current = start
        while current <= end:
            months.append(current.strftime('%b %Y'))
            # Move to the next month
            if current.month == 12:
                current = current.replace(year=current.year + 1, month=1)
            else:
                current = current.replace(month=current.month + 1)

        # Generate random numbers between 20 and 100 for each month
        random_numbers = [random.randint(20, 100) for _ in range(len(months))]
        
        return months, random_numbers




    def convert_to_base64(pil_image):
        buffered = BytesIO()
        pil_image.save(buffered, format="PNG")  # You can change the format if needed
        img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
        return img_str


    def create_graph_1(self):
        # Sample data
        x, y = self.generate_months()
        
        # Create a plot
        plt.figure(figsize=(10, 3))  # Adjust the figure size to fit the left space
        plt.gcf().set_facecolor('#9edbff')
        plt.plot(x, y, marker='s', color='#1b1464')
        for i, label in enumerate(y):
            plt.text(x[i], y[i] - 20, str(label) + "%", ha='center', va='bottom', fontsize=12, color='black')

        plt.ylabel("Helmet Not Use, in %", fontsize=10, weight='bold')
        plt.ylim(0, 100)
        
        # Save the plot as an image
        plt.savefig("./temp/graph_1.png", bbox_inches='tight')  # Save with tight layout
        plt.close()


    def create_graph_2(self):
        data = [30, 20, 24, 26]  # Values for the pie chart
        labels = ['Student', 'Passenger', 'Other', 'Family Member']  # Labels for the slices

        fig, ax = plt.subplots(figsize=(4, 4), facecolor='#9edbff')  # Set canvas background
        #ax.set_facecolor('white')  # Set the pie chart's background to white

        colors = ['#C96868', '#FADFA1', '#FFF4EA', '#7EACB5']

        # Draw the pie chart with labels inside
        wedges, texts, autotexts = ax.pie(
            data, labels=None,  # Exclude outer labels
            colors=colors,
            autopct='%1.1f%%', startangle=90, 
            textprops=dict(color="black", fontsize=12, weight='bold')  # Style the text inside
        )

        # Add original labels to the pie wedges
        for i, autotext in enumerate(autotexts):
            autotext.set_text(f"{labels[i]}\n{autotext.get_text()}")  # Label + percentage

        # Make sure the pie chart is circular
        ax.axis('equal') 

        plt.savefig("./temp/graph_2.png", bbox_inches='tight')  # Save the chart as an image
        plt.close()  # Close the plot to free memory



    def bullet_multicell(self, width, height, bullet, text, bullet_indent=0.4):
        # Add the bullet and first line of text
        self.cell(bullet_indent, height, bullet)  # Bullet with small space
        self.multi_cell(width - bullet_indent, height, text)

        # Adjust X position for wrapped lines (subsequent lines)
        last_y = self.get_y()
        self.set_xy(self.l_margin, last_y)  # Indent wrapped lines
        return last_y


    def generate(self):
        self.set_auto_page_break(auto=True, margin=1.5)
        self.add_page()

        primary_color = (56, 182, 255)
        secondary_color = (27, 20, 100)
        self.image('./resources/banner.png', x=0, y=0, w=21.0)

        self.set_text_color(255, 255, 255)

        self.set_xy(1.0, 1.23)
        self.set_font('Arial', 'B', 24)
        self.cell(0, 0, 'Motorcycle Helmet Safety', ln=False, align='L')
        self.set_xy(1.0, 2.22)
        self.set_font('Arial', '', 18)
        self.cell(0, 0, 'Statistics Report', ln=False, align='L')

        self.set_xy(17.22, 1.73)
        self.set_font('Arial', '', 12)
        self.cell(0, 0.5, 'This document', ln=True, align='L')
        self.set_x(17.22)
        self.cell(0, 0.5, 'was made', ln=True, align='L')
        self.set_x(17.22)
        self.cell(0, 0.5, 'by Knock Knock', ln=True, align='L')

        self.set_draw_color(*primary_color)
        self.set_line_width(0.2)
        self.line(1.0 + 0.2, 3.39 - 0.2, 7.82, 3.39 - 0.2)

        self.set_text_color(0, 0, 0)

        date = f'{self.start_date} - {self.end_date}'

        self.set_xy(1.0, 4.3)
        self.set_font('Arial', '', 14)
        self.cell(0, 0.6, str(uuid.uuid4()), ln=False, align='L')
        self.cell(0, 0.6, date, ln=True, align='R')

        self.ln(0.25)

        self.set_font('Arial', 'b', 21)
        self.cell(0, 1, f'Motorcycle Helmet Non-use', ln=True, align='L')
        self.cell(0, 0.6, f'in {date} Overall Results', ln=True, align='L')

        self.ln(0.75)

        self.set_font('Arial', 'b', 18)
        self.cell(0, 0, 'Camera Information', ln=True, align='L')

        self.ln(0.5)

        self.set_x(1.5)
        self.set_font('Arial', 'b', 14)
        self.multi_cell(19, 0.6, f"{'HP 4M Fixed USB Camera'}", align='L')

        self.set_font('Arial', '', 14)
        
        self.set_x(2.0)
        self.bullet_multicell(19-2, 0.6, "-", f"IP: {'122.22.25.2'}")
        
        self.set_x(2.0)
        #self.multi_cell(19, 0.6, f"- Address: {'1 Chalong Krung 1 Alley, Lat Krabang, Khet Lat Krabang, Krung Thep Maha Nakhon 10520'}", align='J')
        self.bullet_multicell(19-2, 0.6, "-", f"Address: {'1 Chalong Krung 1 Alley, Lat Krabang, Khet Lat Krabang, Krung Thep Maha Nakhon 10520'}")

        self.set_x(2.0)
        self.bullet_multicell(19-2, 0.6, "-", f"Administrator: {'KMITL'}")


        self.ln(0.75)

        self.set_font('Arial', 'b', 18)
        self.cell(0, 0, 'Introduction', ln=True, align='L')
        self.ln(0.5)

        query = f"Write me 2-3 paragraphs about the introduction of Motorcycle Helmet Non-use less than 120 words"
        
        introduction = " "*4 + self.chain.invoke({"text": query, "image": ''}).strip().replace("\t", "").replace("\n", "\n"+" "*4)
        #summary = 3*(' '*4 + "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.\n\n")
        
        self.set_font('Arial', '', 14)
        self.multi_cell(19, 0.6, introduction, align='J')

        self.set_draw_color(*secondary_color)
        self.set_line_width(0.15)
        self.line(1.0 + 0.15, 29 - 0.15, 20, 29 - 0.15)


        self.add_page()


        self.ln(0.25)
        self.set_font('Arial', '', 14)
        self.cell(0, 1.2, 'Figure 1', ln=True, align='L')
        self.set_font('Arial', 'b', 18)
        self.cell(0, 0, f'Motorcycle Helmet Non-use in {date}', ln=True, align='L')

        self.ln(0.4)
        self.create_graph_1()
        self.image("./temp/graph_1.png", x=1, w=19, h=7)

        self.ln(0.25)
        query = f"Analyze the chart depicting the percentage of motorcycle helmet non-use for the month of {date}. Please include key insights such as trends over time, comparisons with previous months or years, any noticeable spikes or drops, and possible reasons for these changes. Summarize your analysis in 1 short paragraph. less than 60 words"
        description_1 = " "*4 + self.chain.invoke({"text": query, "image": "./temp/graph_1.png"}).strip().replace("\t", "").replace("\n", "\n"+" "*4)
        #description_1 = "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum."

        self.set_font('Arial', '', 14)
        self.multi_cell(19, 0.6, description_1, align='J')


        self.ln(0.1)
        self.set_font('Arial', '', 14)
        self.cell(0, 1.2, 'Figure 2', ln=True, align='L')
        self.set_font('Arial', 'b', 18)
        self.cell(0, 0, f'Motorcyclists, by Helmet Type', ln=True, align='L')

        y = self.get_y()

        self.ln(0.4)
        self.create_graph_2()
        self.image("./temp/graph_2.png", x=1, w=9, h=9)

        
        query = f"Analyze the pie chart illustrating the proportion of different categories of individuals who are not wearing motorcycle helmets Discuss Consider potential implications for safety measures or awareness campaigns. Summarize your insights in 1 short paragraph less than 30 words."
        description_2 = " "*4 + self.chain.invoke({"text": query, "image": "./temp/graph_2.png"}).strip().replace("\t", "").replace("\n", "\n"+" "*4)
        #description_2 = "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum."
        
        self.set_font('Arial', '', 14)
        self.set_xy(10.5, y + 0.5)
        self.multi_cell(9.5, 0.6, description_2, align='J')


        self.set_draw_color(*secondary_color)
        self.set_line_width(0.15)
        self.line(1.0 + 0.15, 29 - 0.15, 20, 29 - 0.15)

        filename = f'./temp/motorcycle_helmet_report_{self.start_date}_{self.end_date}.pdf'
        self.output(filename)

        return filename


    def prompt_func(self, data):
        text = data["text"]
        image = data["image"]

        content_parts = []

        text_part = {"type": "text", "text": text}
        content_parts.append(text_part)

        if image != '':
            image_part = {
                "type": "image_url",
                "image_url": f"data:image/jpeg;base64,{image}",
            }
            content_parts.append(image_part)

        return [HumanMessage(content=content_parts)]


    def describe_image(self, text, image_path):
        pil_image = Image.open(image_path)
        image_b64 = self.convert_to_base64(pil_image)

        query_chain = self.chain.invoke(
            {"text": text, "image": image_b64}
        )

        return query_chain


if __name__ == '__main__':
    report = ReportGenerator('NOV 2024', 'NOV 2025')
    #report.generate()
    x, y = report.generate_months()
    print(x, y)
    pass